Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Possible Duplicate:
Can you help me with my capacity planning?

If CPU and RAM on a production host allows more than one JVM (server instances) to run in parallel (to build server farm). How do you arrive at optimal number of JVM processess? What type of "scaling" it is?

share|improve this question

marked as duplicate by Janne Pikkarainen, Chopper3, mgorven, voretaq7 Nov 11 '12 at 8:54

This question was marked as an exact duplicate of an existing question.

up vote 0 down vote accepted

Usually with the JVM you stop when you run out of RAM, or when the heap is squeezed enough that your CPU is thrashed from all the futile garbage collection.

This kinda counts as "scaling out", but you should really be careful. One out of control JVM can and will bring down every JVM on the server. Consider virtualization instead.

share|improve this answer
How? JVMs have maximum heap sizes, so (a) how do they get 'out of control' and (b) how do they bring each other down? – EJP Jul 26 '12 at 6:56
@EJP As I mentioned, CPU thrashing from garbage collection is the main cause that I observe. It can seriously impact other processes when RAM is tight - not just starving other java processes of resources, but the operating system itself. The leap second bug a few weeks back and its effect on JRE processes on Linux systems provided another great example of JVMs going out of control and making the host OS and everything in it unresponsive. – Shane Madden Jul 26 '12 at 16:11

Not the answer you're looking for? Browse other questions tagged or ask your own question.